Fluid Evolution in the Bundelkhand Granite, North Central India: Implications for Hydrothermal Activities in the Bundelkhand Craton

Abstract

The Bundelkhand granite (BG) constitutes the bulk of the granitoid complex in the Bundelkhand Craton and preserves imprints of its evolution from the magmatic to a protracted hydrothermal stage as deduced from the petrography. In order to reconstruct such a path of evolution in this study, thermobarometric calculations were attempted on the mineral chemistry of the major (hornblende, plagioclase, biotite) and minor (epidote, apatite) magmatic phases. They yielded magmatic temperatures and pressures (in excess of 700 °C and ~5 kbar), although not consistently, and indicate mid-crustal conditions at the onset of crystallization. Temperatures in the hydrothermal regime within the BG are better constrained by the chemistry of the chlorite and epidote minerals (340 to 160 °C) that conform with the ranges of homogenization temperatures of aqueous–biphase inclusions in matrix quartz in the BG and subordinate quartz veins. These reconstructions indicate that fluid within the BG evolved down to lower temperatures and towards the deposition of quartz and, more importantly, bears a striking similarity to the temperature–salinity characteristics of fluid in the giant quartz reef system. Scanty mixed aqueous–carbonic inclusions in the BG are indicative of the CO₂-poor nature of the BG magma and the exsolution of CO₂ at lower pressure (~2.6 kbar). The dominant mechanism of fluid evolution in the BG appears to be the incursion of meteoric fluid, which caused fluid dilution. Laser Raman microspectrometry reveals many types of solid phases in aqueous–carbonic inclusions in the BG domain. The occurrence of unusual, effervescent-type inclusions, though infrequent, bears a striking similarity to that reported in the giant quartz reef domain. Thus, the highlight of the present work is the convincing fluid inclusion evidence that genetically links the BG with the giant quartz reef system, although many cited discrepancies arise from the radiometric dates. We visualize the episodic release of silica-transporting fluid to the major fracture system (now occupied by the giant reef) from the BG, thus making the fluid in the two domains virtually indistinguishable.

1. Introduction

Granitoids are the result of crust–mantle interactions at variable scales in diverse tectono-magmatic regimes [1]. Once formed, the granitic magma ascends through the crust before its final arrest within the crust, as dictated by the rheological contrast of the magma and its surroundings [2,3]. The P–T path of evolution of felsic, hydrous magma is expected to be imprinted in mineral phases formed from the early magmatic stages, and later, through the exsolution of the magmatic volatile phase and late-stage hydrothermal activity at lower P–T conditions, manifested by deuteric alteration of the primary minerals and the precipitation of secondary minerals from the hydrothermal fluid as evident from fluid inclusions.

The Bundelkhand craton (BC) occupies the north–central part of the Indian sub-continent, bound by latitudes 24°34′ N to 26°30′ N and longitudes 77°30′ E to 81°51′ E. The BC is largely composed of a granitoid complex in which the pink-to-gray granitoid (or Bundelkhand granite, BG) is the dominant unit, along with subordinate leucogranite and its variants, minor aplitic dykes, and quartz vein networks [4]. The granitoid complex has been studied by various researchers who addressed the geochemical and geochronological aspects [5,6,7,8,9,10,11,12]. However, there is a dearth of information on the evolution of the late-stage fluid, making it difficult to deduce the post-emplacement hydrothermal processes, which are crucial to the origin of the “giant quartz reef” system—a significant post-BG phenomenon [4]. Reconstruction of the evolution of the BG necessitates the retrieval of magmatic, magmatic–hydrothermal transition, and post-magmatic hydrothermal stages from appropriate mineral assemblages, which has been lacking so far. Moreover, there has been no attempt to characterize the late-stage fluid from fluid inclusions. The present study intends to bridge these two important knowledge gaps in regard to the evolution of the BG. Further, we also attempt to examine the genetic link between the BG and the giant quartz reef system using the available information on fluid characteristics in the latter [4].

Numerous studies [5,6,7,8,9,10,11,12] unequivocally establish the geochemical characteristics of the Bundelkhand granitoid to be post-Archaean calc-alkaline, with distinct variations in potassium content, and classify it as meta-aluminous to per-aluminous. Geochronological evidence indicates that the emplacement of the K-rich calc-alkaline Bundelkhand granitoid occurred within a narrow time frame of approximately 30 million years, between 2.57 and 2.54 billion years ago [11]. Chouhan et al. [12] and Sensarma et al. [13] furnished some mineral chemistry data and deduced the P–T conditions of the emplacement of the K-rich Bundelkhand granitoid. The present study focuses on the crystal–melt–fluid system of the BG, emphasizing the mineral chemistry of the constituent minerals that likely formed during different stages of its evolution, along with their textures and associated chemical parameters. This approach aims to clarify the late-stage evolution of fluids within the system. The mineral chemistry of both the primary and secondary phases is utilized to support findings from fluid inclusion studies. This study does not include any bulk chemical characterization of the BG, as these data are already available and not central to the issues being discussed.

2. Geological Background

The Bundelkhand Craton (BC) consists of rocks of the Archaean to Paleoproterozoic age and is overlain by the Late Paleoproterozoic (2.0–1.8 Ga) sedimentary sequence of the Bijawar and Gwalior Groups in the northwestern, south, and southeastern extremities as shown in Figure 1. The Bundelkhand granite is surrounded by the Meso-Neoproterozoic (1.1–1.0 Ga) Vindhyan basin all along the periphery, except on the northern side which is covered by Gangetic alluvium of Quaternary age. The central part of the BC constitutes the Bundelkhand Gneissic Complex (BGC) followed by the Bundelkhand Metasedimentary and Metavolcanic rocks (BMM), Bundelkhand granitoids, and the Madaura (or Madawara) ultramafic complex. The BC, in large part, is dissected by the “giant quartz reef system” comprising 20 major quartz reefs and innumerable minor ones, constituting approximately 10 percent of the area of the craton [4]. The craton is also dissected by mafic dykes that cut across all older litho units, including the giant quartz reefs, as shown in Figure 1.

The exposed Bundelkhand Craton, covering an area of 26,000 square kilometers, is primarily characterized by a granitoid complex, predominantly granite to granodiorite. This granitoid, referred to as the Bundelkhand granite (BG), varies in color from gray to pink and contains differing amounts of pink K-feldspar. Radiometric dating by various workers assigns an age of ~2.5 Ga [6,10,11,14] for BG. Various minor intrusions in the form of felsic units and pegmatitic and quartz veins are also other noticeable features of the BC [11]. Although the BG is the least affected by deformation, in comparison to other older units like the BGC and BMM, brittle shear fractures and ductile shear zones are often observed [13]. Roday et al. [15] visualized the emplacement of the Bundelkhand granitoids as being due to a diapiric rise in a pseudoplastic state, occurring within the dynamic environment of a collision-related volcanic setting [5,6,11,16].

Subordinate to the Bundelkhand granitoids, the quartz reefs constitute the next most voluminous litho units standing as ridges in the BC with a dominant trend of NNE–SSW to NE–SW. Rout et al. [4] surmised a shallow crustal source of the fluid for their formation and speculated it to be the late-stage fluid of the BG that evolved internally within it. Pati et al. [17] reported an age of 1.9–2.0 Ga (K–Ar dating) for the quartz reef, whereas Slabunov et al. [18] inferred a relatively younger age of approximately 1.8 to 1.9 Ga, with the help of the SHRIMP dating of zircons from giant quartz reefs. They obtained a range of dates from 2.8 to 1.2 Ga (U–Pb and Pb–Pb) from these zircons. Notably, these quartz reefs are intersected by NW–SE mafic dykes, demonstrating that the formation of the reefs predates the intrusive mafic dykes [19]. Deb and Bhattacharyya [20] reviewed the regional geology of the Bundelkhand Craton, covering all litho units, and quoted an age range of ~2.0 to 1.1 Ga for the mafic dykes of different generations as determined from radiometric dating.

3. Materials and Methods

Field work was carried out in the BC and representative samples of the BG were collected across the entire body to obtain the maximum possible coverage.

Petrography of the BG was carried out on 50 thin sections in order to understand the primary and secondary textures, representing the pristine magmatic crystallization and the late-stage alteration. Appropriate mineral/mineral assemblages from 15 polished thin sections of the BG were selected for Electron Probe Micro Analysis, which was carried out at two different laboratories. Analyses were carried out with a CAMECA SX FIVE instrument housed at the national facility at the Dept. of Geology (Center of Advanced Study), Institute of Science, Banaras Hindu University (BHU), and a CAMECA SX FIVE instrument at the Central Research Facility, Indian Institute of Technology (Indian School of Mines), Dhanbad, India. A LEICA-EM ACE 200 instrument (Leica Microsystems, Wetzlar, Germany) was used to coat a 20 nm thin layer of carbon on the polished thin section before electron probe micro analyses to accelerate conductivity. Matrix corrections of the raw data were carried out using the ZAF procedure. In both cases, natural mineral standards were used for routine calibration and quantification. The precision of the analysis is better than 1% for major trace element oxides as determined from the repeated analysis of standards.

The main aim of such analyses is to understand the chemistry of minerals constituting the BG in order to trace the physicochemical parameters during the primary magmatic and late hydrothermal stages. Hence, minerals such as plagioclase, K-feldspar, hornblende, biotite, apatite, epidote, garnet, muscovite, titanite, chlorite, and magnetite are analyzed in the current work. Multiple grains of each of the minerals were analyzed, and the data are available in the Supplementary Materials.

Fluid inclusion studies were carried out on doubly polished wafers of the BG and quartz veins in the BG with the thickness of the wafers usually varying from 200 to 300 µm. Fluid inclusion petrography was performed before microthermometry in order to understand their nature, association, and mode of occurrence using a Leica DM 4500 petrological microscope (Leica Microsystems, Wetzlar, Germany). A total of 24 wafers (21 from the BG, 3 from quartz veins in the BG) were used for the fluid inclusion study. Fluid inclusions were studied in the matrix quartz in the BG and also in the late-stage quartz veins/stringers which form as a result of late-stage fluid activity in the BG. The population of fluid inclusions and the nature of their distribution in wafers were deciphered by observing them at varying magnifications. Fluid inclusion microthermometry was carried out at the Indian Institute of Technology, Kharagpur, with the help of a Linkam THMSG 600 hot–cold stage attached to a Leica DM 2500 petrological microscope with a 50× long-distance working objective. The experimental details of the fluid inclusion study are as detailed in [4]. The interpretations of freezing data for aqueous inclusions were made with the help of standard equations for the freezing curve of the H₂O–NaCl system, whereas the standard P–V–T–X relationships in H₂O–NaCl, H₂O–CO₂–NaCl, and CO₂–CH₄ systems were used for the computation of fluid densities and the construction of isochores. All calculations were carried out using an MS Excel–MS Visual Basic Macro based program [21].

Raman microspectrometric studies were carried out at the Research School of Earth Sciences, Australian National University. A Renishaw inVia Reflex Spectrometer System equipped with a standard confocal microscope was used for Raman spectral analysis. A Renishaw diode-pumped solid-state laser provided 532 nm laser excitation with 5 mW power to the sample. A 2400 grooves/mm grating was used, giving a spectral resolution of 0.5 cm⁻¹. Single Raman spectra were obtained using a 5 s integration time with 10 accumulations and a 100× Leica microscope objective, which focused the beam to a spot size of 1 µm. Wave number calibration was carried out using an internal silicon standard and was performed as an automated procedure using the Wire version 4.2 software.

Two-dimensional Raman mapping was carried out using the StreamLineHR™ mode of the Wire software. The maps were produced by rastering the region of interest in 5 µm steps and recording a Raman spectrum at each step with a 1 s integration time and 1 accumulation. Maps of different minerals were produced by plotting the intensity of definitive Raman bands for each mineral at each step. These were then displayed as colored maps using the Wire software version 4.2.

4. Field Features

Small-to-large outcrops of the BG are uniformly distributed all over the craton, showing variable degrees of alteration. They are leucocratic to mesocratic with an inequigranular and holocrystalline nature. The pink variant of the BG is prevalent with occasional occurrences of the gray type. The BG is dominantly coarse-grained and shows a hypidiomorphic texture. K-feldspar (Figure 2a) and quartz of variable grain sizes are commonly seen as megacrysts in them. Mafic minerals occur as fine to medium grained, sometimes as coarse or very coarse-grained in nature. Patches of epidote are commonly observed.

The BG is devoid of any mesoscopic structure. However, alternate bands of felsic and mafic minerals adjacent to the Bundelkhand Tectonic Zone (BTZ) occasionally show poor to well-developed NE–SW foliation (Figure 2b). In addition to slip surfaces, mylonitic gray granite in a quarry section close to Kabrai (north–eastern part) exhibits sigmoidal porphyroclasts showing a sense of shear from east to west with slip lineation N85° dipping 87° towards north. Occasional grains of sulfide minerals (molybdenite?) are observed in the BG in close proximity to the BTZ. Sparse copper sulfide (mostly oxidized on the surface) and molybdenite are observed in the BG near the Raksha Shear Zone (RSZ) area as reported by Pati et al. [9]. Aplite veins are also seen in the BG (Figure 2c). The presence of pegmatites and quartz to quartzo-feldspathic veins indicates late-stage fluid activity in the BG (Figure 2d,e) [22,23]. Often, minor faults are seen to displace such veins as well as zones of ductile shearing associated with the BG (Figure 2d). Occasionally, parallel joints or fault planes are also observed in the BG. The occurrence of mafic enclaves is quite common in the BG. Granite outcrops are better observed in the proximity of the resistant quartz reefs (Figure 2f). In such cases, they are significantly altered, friable, and foliated (Figure 2g).

5. Petrography

The Bundelkhand granite (BG) exhibits an inequigranular texture, with granularity ranging from medium to coarse-grained. It primarily consists of quartz, plagioclase feldspar, and K-feldspar, with a notable presence of hornblende and biotite. Additionally, there is less than a 1% modal percentage of accessory phases which include epidote, chlorite, titanite, apatite, allanite, and zircon. Restricted occurrences of bleached mica and garnet are seen in a few samples. Both iron oxide and iron sulfide appear as fine-to-medium-grained opaque minerals with varying habits.

Subhedral-to-euhedral plagioclase is typically unzoned. It primarily occurs as megacrysts and exhibits varying degrees of saussuritization. Patches of quartz, biotite, clinozoisite, and chlorite are often observed to be in close association with plagioclase. Unaltered grains retain their lamellar twinning. K-feldspar is dominantly microcline, with perthitic intergrowth (Figure 3a) observable in the forms of patchy perthite or flame perthite. In contrast to the fresh appearance of the host K-feldspar, the plagioclase within these perthitic intergrowths appears turbid (finely altered), which is more akin to patchy perthite due to deuteric alteration [24,25]. In some instances, both plagioclase and K-feldspar grains exhibit kinks and criss-cross fractures, with the fracture spaces filled by fine-grained quartzo-feldspathic materials and other secondary phases, viz. epidote, chlorite, and sericite. A narrow zone of intense granulation of K-feldspar resulting in a mineral fish-like structure is often observed, which suggest at least one episode of shearing in the BG. The signs of dissolution and precipitation can be marked in the thin sections at variable scales with serrated margins and a subgrain formation of quartz.

Quartz occurs as anhedral aggregates that exhibit sweeping or undulose extinction. It is mostly interstitial to the early-formed mineral phases and is typically coarse-grained except when it appears as veins that traverse the matrix. Serrated margins, grain boundary migration, and subgrain formation (Figure 3b) are some of the notable features in them depicting dissolution–reprecipitation and a recrystallization signature.

Hornblende is commonly coarse-grained indicating its early crystallization, whereas biotite displays variable grain size, depicting both a primary as well as a secondary origin. Hornblende displays pleochroism from light green to dark green (Figure 3c) as well as from straw yellow to dark green—the various pleochroic shades may be ascribed to variations in Ti content [26]. Biotite also displays variable degrees of pleochroism such as faint brown to dark brown or dark green, and sometimes exhibits combinations of these colors. The presence of biotite along the fringe or cleavage plane of hornblende depicts retrogression reactions of hornblende in the presence of fluid (Figure 3d). Such biotite grains are relatively smaller in size and are devoid of any deformation, whereas the coarser grains show signs of deformation in terms of kinks/bent cleavage planes and bird’s eye extinction (Figure 3e). Incipient-to-complete chloritization in the form of a pseudomorph of biotite is observed with a gray and purplish interference color (Figure 3e) which is sometimes associated with needles of rutile (release of Ti) and fluorite (release of F) in the cleavage spaces and also secondary titanite. In a couple of samples, biotite of an anhedral-to-subhedral habit (30 to 800 μm) is observed to have been completely bleached to a colorless appearance. While it lacks distinct bodily pleochroism, a faint brownish hue and strong relief are visible along its borders and cleavage planes, as shown in Figure 3f. Additionally, it exhibits a second-order interference color similar to that of white mica. Occasionally, it shows kinks over cleavage planes and displays cat’s eye extinction with fine-grained aggregates of white mica surrounding its edges.

There is primary as well as secondary titanite. Titanite along with epidote and chlorite are the breakdown/alteration products of both hornblende and biotite (Figure 3c,d). Apatite mostly occurs in its typical euhedral habit as basal, oblique, or prismatic sections. Apatite grains often display a fractured nature. Zircon, with its primary magmatic oscillatory zoning, is also observed as inclusions within biotite/hornblende as well as within the quartzo-feldspathic matrix. Magnetite occurs as a common alteration product in the form of fine specks in hornblende (Figure 3g) and other mafic assemblages as seen in Figure 3h. It is also seen at the core of titanite grains. Euhedral-to-subhedral grains of magnetite (Figure 4a) frequently display martitization, characterized by fine lines and irregular patches. Such magnetite grains are associated with disseminated fine-grained iron sulfide. Additionally, subhedral-to-euhedral grains of pyrite can be observed along with fine-grained magnetite. Pyrite rimmed by goethite and rare chalcopyrite are also found dispersed within the quartzo-feldspathic matrix. The presence of goethite, along with a network of reddish-brown stains, indicates subsequent fluid activity.

Epidote is found in various mineral assemblages, displaying different textural characteristics that indicate both magmatic and hydrothermal origins. It is often zoned with an allanite core (Figure 4b). Epidote is more frequently observed to be completely or partially enclosed in biotite (Figure 4c) showing a resorbed margin. Epidote, exhibiting resorbed margins of variable extent, is often found in association with the quartzo-feldspathic matrix (Figure 4d,e). Micro-veins containing fine-grained epidote are definitive indications of late hydrothermal fluid activity.

Garnet is faint brown in color, sub-rounded to crudely hexagonal in habit with a size varying from 50 to 100 µm (Figure 4f). It is unzoned and commonly fractured. Fine aggregates of mica showing barely any pleochroism are found along the garnet edges without any chloritic alteration. In some of them, quartz inclusions of 5–20 µm are also seen. These are restricted in occurrence.

6. Mineral Chemistry

Plagioclase feldspar in the BG is mostly albitic to oligoclase in nature with X_Ab varying from 0.73 to 0.99, while X_Or varies from 0.002 to 0.06 and X_An varies from 0.005 to 0.26. Alkali feldspar furnished an X_Ab value of 0.02 to 0.1 and X_Or values of 0.89 to 0.97. The anorthitic component in the alkali feldspar is negligible and varies from 0 to 0.002 (Figure 5a).

Hornblende is of the calcic amphibole type with Na values 0.0 to 0.44 apfu and Ca values in the range of 1.31 to 1.95 apfu (on a 22-oxygen basis). They show both magnesian and ferroan affinity with a broad variation in Mg# from 0.24 to 0.64 with a dominant spread over 0.48 to 0.58 as seen in Figure 5b. The total aluminum content in them varies from 0.80 to 2.01 apfu with an Al ^IV and Al ^VI occupancy of 0.020130.42 and 0.61–1.96 apfu, respectively. Almost all the hornblende analyses indicate an igneous origin as shown in Figure 5c [27]. All the calculations of hornblende in the current work were performed using the method of Yavuz [28].

Biotite has undergone variable degrees of chloritization. Thus, the data were filtered out according to the compositional range given by Deer et al. [29] and were interpreted with the help of Mica+ [30]. The analyzed biotite spots (n = 32) are plotted on the classification diagram of [31] and furnished a range in Al ^IV and Al ^VI from 1.09 to 1.23 apfu and 0.55 to 0.60 apfu, respectively, with a broad Mg# value of 0.34 to 0.64 (Figure 5d). Biotite from the BG is all re-equilibrated primary biotite (Figure 5e) [32], and the data are distributed in fields II and III of the diagram by de Albuquerque [33], which means it is neither accompanied by other ferromagnesian minerals nor coexisting with muscovite (Figure 5f). It mostly shows a calc-alkaline affinity of orogenic suites [34] with minor deviations.

Figure 5. (a) Distribution of feldspar analyses in a ternary diagram showing the fields of plagioclase series and alkali feldspar [29]; (b) the classification diagram of calcic amphibole [28]; (c) a scatter plot showing the igneous origin of hornblende [27]; (d) the classification diagram of biotite [31]; (e) biotite showing the re-equilibrated primary nature in the diagram of FeO_total–TiO₂–MgO [32]; (f) a ternary plot (Al₂O₃–FeO_tot–MgO) of mica samples [33] where I represents biotite coexisting with amphibole, II represents biotite unaccompanied by other ferromagnesian minerals, III represents biotite coexisting with muscovite, and IV represents biotite coexisting with aluminosilicates.

Figure 5. (a) Distribution of feldspar analyses in a ternary diagram showing the fields of plagioclase series and alkali feldspar [29]; (b) the classification diagram of calcic amphibole [28]; (c) a scatter plot showing the igneous origin of hornblende [27]; (d) the classification diagram of biotite [31]; (e) biotite showing the re-equilibrated primary nature in the diagram of FeO_total–TiO₂–MgO [32]; (f) a ternary plot (Al₂O₃–FeO_tot–MgO) of mica samples [33] where I represents biotite coexisting with amphibole, II represents biotite unaccompanied by other ferromagnesian minerals, III represents biotite coexisting with muscovite, and IV represents biotite coexisting with aluminosilicates.

Structural formulae of bleached mica were calculated on the basis of 11 anhydrous oxygens using the Excel sheet of Verdecchia et al. [35]. The apfu value of Si in bleached mica varies from 2.93 to 3.24 with an Al and total alkali content of 2.40–2.87 and 0.78–0.97, respectively. The obtained X_Mg (i.e., Mg/Mg + Fe) values vary between 0.09 and 0.46. According to Speer [31], the data points show compositional variations from primary (I) to late-post magmatic (II) type (Figure 6a). The variation in the chemical composition can be marked with the substitution vector as indicated by the arrow in Figure 6b,c.

The stoichiometry calculation of apatite was conducted with reference to Ketcham [36] as per 25 oxygen equivalents. Apatite in the BG has CaO values in the range of 51.29 to 57.89 wt% and P₂O₅ values from 39.15 to 43.46 wt%. The F content varies from 2.4 to 6.4 wt%, whereas Cl ranges from 0.0 to 0.34 wt%. According to Piccoli and Candela [37], apatite with an F content up to 3.76 wt% is generally fluor-apatite, which is most commonly seen in apatite of igneous origin. The MnO concentration in such apatite varies from 0.0 to 0.23 (except for one value showing 0.33 wt%). The apatite data mostly suggest it is of the non-ore bearing type [37], although both fields overlap as shown in Figure 6d.

The crystal-chemical formula of titanite (based on 20 oxygens) suggests Ca ranges from 3.77 to 4.01 apfu, Al = 0.18 to 0.37 apfu, Fe = 0.13 to 0.23 apfu, and Ti = 3.51 to 3.86 apfu. All of the spot analyses show X_Al = Al/(Al + Ti + Fe³⁺) < 0.25, indicating that it is not high-aluminum titanite [38].

Figure 6. (a) Compositional variations in muscovite in the TiO₂–FeO–MgO diagram [31], where the abbreviations suggest the following: I—Primary; II—Late-Post magmatic; and III—Hydrothermal muscovite. (b,c) Chemical composition of white mica in the BG. Arrows suggest different substitution vectors [35]. (d) Apatite classification diagram [37]. (e,f) Chlorite classification diagram calculated with WinCcac [39].

Figure 6. (a) Compositional variations in muscovite in the TiO₂–FeO–MgO diagram [31], where the abbreviations suggest the following: I—Primary; II—Late-Post magmatic; and III—Hydrothermal muscovite. (b,c) Chemical composition of white mica in the BG. Arrows suggest different substitution vectors [35]. (d) Apatite classification diagram [37]. (e,f) Chlorite classification diagram calculated with WinCcac [39].

The chlorite data were retrieved based on 14 oxygens using the WinCcac program [39]. The analyzed chlorites are dominantly classified as chamosite followed by some clinochlore [40] (Figure 6e,f). The total tetrahedral occupancy (apfu) is found to be full, i.e., 4, whereas it varies from 5.8 to 5.9 with very few analyses showing full occupancy for the octahedral sites. Some of the chlorite analyses were found to be contaminated by their host which is supported by Na₂O + K₂O + CaO > 0.5 wt% [35]. Such data points were discarded while interpreting the chlorite dataset. The rest of the analysis suggests Si values of 2.5 to 3.0 apfu and Al values of 2.0 to 2.9, and X_Mg = Mg/(Mg + Fe) is 0.25–0.65.

Epidote usually forms a solid solution between pistacite and clinozoisite. In the present case, the pistacite content [Ps = Fe³⁺/(Fe³⁺ + Al)] shows a broad range from 12 to 33 with TiO₂ = 0.000 to 0.338 wt%. Sensarma et al. [13] reported a higher Ps value of 0.24 to 0.37 in the BG and reported it to be of magmatic origin. An attempt has been made to understand the broad variation in Ps content in the current work by segregating it into each group based on texture and the mineral assemblages as discussed in the petrography section. The Ps content in subhedral-to-anhedral epidote which is partially or completely enclosed by a biotite grain varies from 0.26 to 0.33 (n = 15; except one, 0.20) whereas resorbed epidote with a subhedral-to-anhedral nature associated with plagioclase feldspar furnished a Ps mole fraction of 0.19 to 0.29 (n = 9). The prismatic epidote associated with quartz furnished a narrow variation in Ps content, i.e., 0.26 to 0.29 (n = 4), whereas a broader value of 0.21 to 0.30 (n = 14, except one Ps = 0.12) was observed for epidote associated with iron sulfide minerals. The Ps content in epidote veins varies from 0.19 to 0.27 (n = 9). Although textural study suggests the magmatic nature of some epidote groups, the broad range in Ps content adds some equivocal constraints. For a magmatic epidote, it is highly desirable that the Ps content be between 0.25 and 0.29 [41] and the TiO₂ value be less than 0.2 weight percent [42]. The above chemical constraints in terms of Ps and TiO₂ were also supported by Sial et al. [43] and Galindo et al. [44] in their study of high-K calc-alkaline plutons. However, Sial et al. [45] reported some broad values (Ps = 0.21 to 0.29) for high-K calc-alkaline pluton. They also observed variation in Ps content for different textured epidote. Moreover, Zen and Hammarstrom [46] and Brasilino et al. [47] give emphasis to some textural criteria while distinguishing magmatic epidotes from secondary epidotes. They are as follows: euhedral-to-subhedral habit, twinning and strong compositional zoning, allanite-rich core, crystallization after hornblende and before or contemporaneous with biotite, and in some cases embayed, where in contact with feldspar grain, wormy contacts (almost myrmekititic), a lack of alteration of biotite to chlorite, and the fresh appearance of plagioclase.

A total of 37 spot analyses were performed on different grains of garnet restricted to only two samples of the BG. The mineral chemistry data are interpreted in terms of pyrope–almandine–spessartine solid-solution-end members of garnet. They vary from 2.15 to 3.42, 48.53 to 54.99, and 41.88 to 48.51 mol%, respectively. Garnet in this study has a reported very small fraction of grossular content which ranges from 0.36 to 1.21 mol%. According to [48], the normal variation in spessartine content in garnet from granite is 36%, whereas it is 47% in garnet from Pegmatites. Hence, the high proportion of spessartine in the present case also points to the typical nature of granitic rocks.

The analyzed iron oxides are found to be mostly magnetite followed by ilmenite with some titano-magnetite. The FeO content varies from 38.74 to 93.86 wt% with TiO₂ and MnO contents in the range of 0.0 to 57.73 wt% and 0.0 to 9.75 wt%, respectively. None of the magnetite analyses totaled 100 wt% even after repeated analyses, whereas ilmenite did. Similar issues were also encountered in the dataset provided by [13]. In their analysis, FeO content varied from 39.01 to 95.24 wt% with TiO₂ and MnO contents in the range of 0.0 to 47.90 wt% and 0.0 to 3.46 wt%, respectively. According to Smith [49], the analyzed magnetite is a solid solution between magnetite and maghemite, which in turn is unable to accommodate the excess Fe₂O₃ component in it. This solid-solution reaction is an indication of sub-solidus oxidation [50]. It is observed that some of the analyzed ilmenite grains showed a relatively higher value of MnO in comparison to others, i.e., 1.0 to 9.75 wt%. This could be due to the formation of a solid solution of ilmenite with geikielite (MgTiO₃) and pyrophanite (MnTiO₃), which indicate that these minerals can substitute for each other in the crystal structure.

7. Results: Retrieval of Physicochemical Environment

7.1. Mineral Thermobarometry

Pairs of coexisting plagioclase and alkali feldspar in the BG were used for the estimation of temperature following the method of Benisek et al. [51]. The range of pressure as estimated from the Al in the hornblende barometer of Schmidt [52] was used here. The results show a variation of 456.0 to 642 °C (avg = 525.7 °C; ±σ = 68.7 °C) and 567.9 to 813.7 °C (avg = 664.6 °C; ±σ = 87.2 °C) for the albite and orthoclase components in them, respectively. The non-overlapping nature of such data (Figure 7a) indicates their retrograde resetting in response to slow cooling. This is also corroborated by the relatively lower value of the orthoclase component in plagioclase compared to that of the albite content in K-feldspar.

Pressures estimated from the composition of hornblende from the BG are tabulated in Table 1 using formulations suggested by various workers. Although the sample variations are almost similar in all the formulations for the BG, the equation of Schmidt [52] furnished a relatively higher value in many instances. A broad variation in emplacement pressure from 0.56 to 5.9 kbar was observed for the BG by using the formulations of Johnson and Rutherford [53] and from 1.48 to 5.97 kbar by using Schmidt [52], respectively, whereas a pressure value of 0.61 to 6.05 kbar is furnished from Molina et al. [54]. However, the stabilization of hornblende would require a minimum of >~1 kbar pressure in a typical granitic system [55]. Hence, the output from Schmidt’s [52] formulation is more acceptable than that of the others’ formulations. These pressure values were considered as per the specifications given by Anderson and Smith [56] in the context of the control of temperature and magma composition over pressure as suggested by Putrika [57]. Furthermore, the machine-learning approach of Li and Zhang [58] for biotite-bearing magma (n = 14) furnished a comparatively narrow range of pressure values of 3.92 to 5.32 kbar corresponding to temperatures of 736 to 770 °C. In comparison to this, Sensarma et al. [13] reported a higher pressure range of 5 to 9 kbar for those samples with T > 800 °C and 6 to 7.5 kbar for T < 800 °C. A few of the analyses from the southern part of the terrane also furnished lower pressure values of ~2 to 3 kbar.

Ti in hornblende thermometry [26] has been used where titanite or ilmenite occurs as an equilibrium mineral phase that furnishes a temperature range from 577 °C to 727 °C with an average of 654 ± 50.4 °C, whereas the Ti in hornblende thermometry by Liao et al. [59] resulted in temperature values from 472 °C to 808 °C with an average of 649 ± 68 °C. However, the later formulation is used for subalkaline systems. Plagioclase–hornblende thermometry on the basis of Al^IV-A site coupling in amphibole [60] has also been attempted. Before using the thermometer, the dataset was checked with the calibration values provided by Holland and Blundy [61]. Since the formulation is pressure-dependent, the p-values calculated with the help of Schmidt [52] as well as Anderson and Smith [56] were considered as a starting point during each iteration. However, most of the data have [Fe³⁺/(Fe³⁺ + Fe²⁺)] < 0.25, which is outside the calibration range given by Anderson and Smith [56]. Hence, very few calculations could be performed using this method. All the calculations are conducted with a Microsoft excel spreadsheet “Plagioclase-Hornblende-Thermobarometry” by Anderson et al. [62] and the resultant temperature varies from 523.6 °C to 724.2 °C with an average of 650.2 °C (±1σ = 54.8 °C) which is within the validation field of Holland and Blundy [61].

Table 1. Sample-wise variation in estimated pressure and temperature after hornblende mineral analysis (EPMA) with the help of various formulations such as P₁: Hammerstrom and Zen [63]; P₂: Johnson and Rutherford [53]; P₃: Hollister et al. [64]; P₄: Schmidt [52]; T₁ and T₂: plagioclase and hornblende thermometer of Holland and Blundy [61] by utilizing pressure values of Schmidt [52] and Anderson and Smith [56], respectively; P₅: Molina et al. [54]; T₃: titanium concentration in hornblende by Otten [26]; T₄: titanium concentration in hornblende by Liao et al. [59].

Sample No.		D1-5i (n = 7)		D2-4i (n = 1)		D2-7 (n = 6)		D4-3 (n = 11)		D2-5 (n = 8)		D2-12i (n = 11)		D6-16B (n = 29)
	P–T Values
		Min	Max	Min	Max	Min	Max	Min	Max	Min	Max	Min	Max	Min	Max
P1 (±3 kbar)		0.83	3.81	0.97	-	4.95	5.18	1.51	3.78	1.25	2.87	1.87	4.16	0.47	5.57
P2 (±0.5 kbar)		0.56	3.91	0.72	-	5.19	5.45	1.33	3.88	1.04	2.86	1.74	4.29	0.16	5.88
P3 (±1 kbar)		0.53	3.04	0.65	-	4	4.2	1.11	3.02	0.89	2.25	1.41	3.33	0.23	4.52
P4 (±0.6 kbar)		1.48	4.31	1.61	-	5.38	5.61	2.13	4.28	1.89	3.42	2.47	4.63	1.14	5.97
P5 (±1.5 to ±2.3 kbar; expressed at 1 s)		0.61	-	1.82	3.17	4.77	6.05	0.96	2.69	1.97	3.84	3.6	-	1.17	-
Plagioclase–Hornblende Thermometer
T1(°C)		536	651	-	-	666	712	628	686	524	553	659	672	-	-
T2 (°C)		-	-	-	-	-	-	674	686	-	-	-	-	-	-
Titanium in Hornblende Thermometer
T3 (°C)		582.3	727	592	669	647	724	583	727	577	635	620	623	581	690
T4 (°C)		516	751	549	695	644	751	516	738	472	808	577	751	516	710

Table 1. Sample-wise variation in estimated pressure and temperature after hornblende mineral analysis (EPMA) with the help of various formulations such as P₁: Hammerstrom and Zen [63]; P₂: Johnson and Rutherford [53]; P₃: Hollister et al. [64]; P₄: Schmidt [52]; T₁ and T₂: plagioclase and hornblende thermometer of Holland and Blundy [61] by utilizing pressure values of Schmidt [52] and Anderson and Smith [56], respectively; P₅: Molina et al. [54]; T₃: titanium concentration in hornblende by Otten [26]; T₄: titanium concentration in hornblende by Liao et al. [59].

Sample No.		D1-5i (n = 7)		D2-4i (n = 1)		D2-7 (n = 6)		D4-3 (n = 11)		D2-5 (n = 8)		D2-12i (n = 11)		D6-16B (n = 29)
	P–T Values
		Min	Max	Min	Max	Min	Max	Min	Max	Min	Max	Min	Max	Min	Max
P1 (±3 kbar)		0.83	3.81	0.97	-	4.95	5.18	1.51	3.78	1.25	2.87	1.87	4.16	0.47	5.57
P2 (±0.5 kbar)		0.56	3.91	0.72	-	5.19	5.45	1.33	3.88	1.04	2.86	1.74	4.29	0.16	5.88
P3 (±1 kbar)		0.53	3.04	0.65	-	4	4.2	1.11	3.02	0.89	2.25	1.41	3.33	0.23	4.52
P4 (±0.6 kbar)		1.48	4.31	1.61	-	5.38	5.61	2.13	4.28	1.89	3.42	2.47	4.63	1.14	5.97
P5 (±1.5 to ±2.3 kbar; expressed at 1 s)		0.61	-	1.82	3.17	4.77	6.05	0.96	2.69	1.97	3.84	3.6	-	1.17	-
Plagioclase–Hornblende Thermometer
T1(°C)		536	651	-	-	666	712	628	686	524	553	659	672	-	-
T2 (°C)		-	-	-	-	-	-	674	686	-	-	-	-	-	-
Titanium in Hornblende Thermometer
T3 (°C)		582.3	727	592	669	647	724	583	727	577	635	620	623	581	690
T4 (°C)		516	751	549	695	644	751	516	738	472	808	577	751	516	710

According to Henry et al. [65], the Ti content in biotite can be used to estimate the solidus temperature of the rock in which it occurs provided it should have X_Mg = 0.275–1.000 and Ti = 0.04–0.60 apfu. In view of the variable degree of chloritization, biotite is expected to lose Ti and Ba on its rims and along the bands parallel to cleavage planes with the advancement of sub-solidus cooling. Hence, the analyzed biotite grains were crosschecked with the aforementioned compositional range before accepting the calculations. The calculated temperature values vary from 489 to 669 °C, which is well within the calibration range provided by them, i.e., 480 to 800 °C. These data are nearly conformable with the temperature range inferred from the re-equilibrated biotite of Sensarma et al. [13], which is 550 °C to 600 °C. However, they have reported a temperature value of ~750 °C for primary biotite. Most of the biotite in our study plotted between the contours of 600 °C–650 °C in the Ti (apfu) vs. Mg/(Mg + Fe) diagram (Figure 7b) with very few points near the 500 °C isotherm. The precision of this geothermometer is about ±24 °C at the lower range and ±12 °C for the higher temperature range. Additionally, the co-occurrence of biotite and apatite is also considered for the estimation of temperature on the basis of F-OH exchange [66]. Although such occurrences are very common in the BG, variable degrees of chloritization will always introduce an unknown degree of uncertainty. Moreover, the analyzed apatite grains gave a more than 2 apfu range for the F content in most of the analysis. Thus, ambiguity arises as to whether they are the result of a substitution reaction and/or the effect of F migration towards the beam [67]. Hence, they were discarded from further calculations, which resulted in very few data points (n = 5) that are amenable to apatite–biotite thermometry and are restricted to only one sample. The calculated temperature varies broadly from ~394 to 824 °C with an average of 612.9 ± 170.6 °C. The Ti–biotite thermometer in the same sample furnished a narrow temperature range from ~629 °C–669 °C (n = 8). The formulation which is used here is pressure-dependent, and thus the pressure values calculated from the Al content in hornblende [52] are utilized at such junctures.

Utilizing the solid solution model of Bird and Helgeson [68], an attempt has been made to calculate temperatures from epidote compositions on the basis of Al-Fe exchange in the octahedral site. A broad range from ~161 °C to 573 °C (Figure 7c) is obtained using the above formulation (calibration value of 5 kbar at 600 °C). All of these data points are within the TiO₂ value < 0.2 wt%. Epidote, which is fully or partially enclosed within biotite, suggests a temperature range from 450 °C to 573 °C and that associated with plagioclase with a resorbed outline suggests a temperature interval of 395 °C to 455 °C. The precise details are given in

Table 2. Temperature variation in epidote with different textures and mineral associations using the solid-solution model of Bird and Helgeson [68] along with the corresponding pistacite variation in them.

Epidote + Mineral Assemblage	Texture of Epidote	Pistacite Content	Temperature (°C)	Remark
Biotite	Fully enclosed grains (n = 13)	0.26–0.33	450.2–573.6 (one 334.1)	Ps = 0.26 → T = 334.1 °C Ps = 0.27–0.29 → T = 450.2–547.7 °C Ps = 0.29–0.33 → T = 471.6–573.6 °C
	Partially enclosed grains (n = 3)	0.26–0.27 (one Ps = 0.20)	441.7–497.9 (one 286.9)	Ps = 0.20 → T = 287 °C Ps = 0.26–0.27→T = 441.7 °C–497.9 °C
Plagioclase	Subhedral and resorbed outline (n = 2)	0.19 and 0.29	455 and 603.8	Ps = 0.29 → T = 603.8 °C is outside the calibration value
	Irregular type (n = 7)	0.23–0.27	328.4- 456.2
Quartz	Elongated prismatic crystals (n = 4)	0.26–0.29	355.8–416.8
Opaque	Fe sulfide (n = 14)	0.21–0.30 (one Ps = 0.12)	332.4–474 (one T = 208.7)	Ps = 0.12 → T = 208.7 °C Ps = 0.21 → T = 211.9 °C Ps = 0.23–0.25 → T = 297.6–337 °C Ps = 0.26–0.30 → T = 365.1–441.4 °C
Plagioclase or quartz-feldspar	Solution channel or fracture-filling material (n = 9)	0.19–0.27	161.8 to 366	Ps <0.25 (0.19–0.24) → T = 161.8 –303.5 °C (n = 7) Ps = 0.25–0.29 → T = 317–366 °C (n = 2)

. No epidote analyses were available in association with the analyzed biotite for a comparative study.

Considering the variations in the mineral composition of chlorite due to the multi-component solid solution, many approaches have been explored to develop a suitable geothermometer. The temperature data as calculated by using the empirical formulations [69,70] displayed a consistent variation. The overall range is from 268 to 337 °C. In contrast to the above, the empirical formulation as proposed by Xie et al. [71] furnished a broader range of temperature of ~167 to 475 °C with a clustering of data points mostly around 200 to 400 °C (Figure 7d). In case of the semi-empirical thermometer with Fe_Total = FeO [72,73,74], the temperature values varied from 217 to 340 °C, except one sample with comparatively lower temperature values, viz. 143 to 172 °C. Thus, the temperature data as obtained from the empirical formulations [69,70] and the semi-empirical thermometer with Fe_Total = FeO [72,73,74] are well correlated and also come under the calibration range provided by respective workers (see

Table 3. Sample-wise variation in estimated temperature after chlorite mineral analyses in the BG. T₁: Innoue et al. [74]; T₂: Lanari et al. [73]; T₃: Bourdelle et al. [72]; T₄: Cathelineau [70]; T₅: Jowett [69].

Sample No.		Semi Empirical Thermometry (FeTotal = FeO)	T1 FeTotal = Feo (Valid for <350 °C, May Be <400 °C)	T2 FeTotal = Feo (Valid for 100–500 °C)	T3 FeTotal = Feo (Valid for <350 °C)	Empirical Thermometry (FeTotal = FeO)	T4 (Valid for <350 °C)	T5 For Si < 3.3,Ca < 0.07 apfu, Fe/(Fe + Mg) < 0.6; valid for T < 325 °C)
D1-5	Av		261.5 (n = 2)	-	-		302 (n = 2)	308.5 (n = 2)
	±1σ		2.12	-	-		2.82	3.53
D2-4	Av		307 (n = 1)	-	-
	±1σ		-	-	-		-	-
D2-12	Av		268.7 (n = 8)	225.6 (n = 3)	306.7(n = 7)		281.7 (n = 8)	281.6 (n = 7)
	±1σ		18.09	47.75	21.23		26.1	20.95
D4-1	Av		264.4 (n = 7)	217.0 (n = 1)	285.5 (n = 2)		275.7 (n = 7)	280.2 (n = 7)
	±1σ		19.51	-	38.89		14.56	15.3
D4-3	Av		235.0 (n = 1)	253.0 (n = 1)	292.0 (n = 1)		300.0 (n = 1)	302.0 (n = 1)
	±1σ		-	-	-		-	-
D5-7	Av		250.0 (n = 3)	251.5 (n = 2)	267.0 (n = 2)		328.0 (n = 2)	-
	±1σ		33.15	19.09	12.72		-	-
BQ-D2-5	Av		252.3 (n = 3)	-	312.0 (n = 2)		281.0 (n = 3)	287.0 (n = 3)
	±1σ		49.54	-	42.43		19.28	19.29
D6-16B	Av		257.3 (n = 13)	258.0 (n = 1)	245.0 (n = 4)		268.0 (n = 13)	273.6 (n = 13)
	±1σ		44.22	-	46.23		15.43	15.74
D6-13	Av		289.2 (n = 4)	305.3 (n = 3)	340.5 (n = 2)		337.0 (n = 1)	-
	±1σ		9.87	20.59	2.12		-	-
D6-16A	Av		281.8 (n = 14)	295.5 (n = 9)	300.67 (n = 6)		325 (n = 5)	311.0 (n = 2) -
	±1σ		26.4	62.12	33.88		20.78	7.07
D6-14	Av		279.2 (n = 5)	232.5 (n = 2)	-		300.0 (n = 2)	313.0 (n = 2)
	±1σ		45.55	13.43	-		8.48	8.48
BAMH-G1	Av		172 (n = 1)	143.0 (n = 1)	171 (n = 1)		295 (n = 1)	307 (n = 1)
	±1 σ			-	-
D1-1	Av		293.1 (n = 7)	236.0 (n = 3)	317.6 (n = 3)		296.0 (n = 7)	299.5 (n = 6)
	±1σ		26.76	17.69	21.93		17.34	11.94
D1-3	Av		253.8 (n = 7)	249.7 (n = 4)	280.5 (n = 3)		309.0 (n = 7)	300.8 (n = 4)
	±1σ		27.3	41.59	38.54		26.19	15.53

). All the calculations were conducted using the Excel sheet of Verdecchia et al. [35].

Since the chlorite in the BG is undoubtedly of late-hydrothermal origin (both as alteration of preexisting biotite/amphibole and as vein precipitates/patches), the temperature estimations derived from its mineral chemistry may also provide some information regarding the thermal regime of the late-stage fluid. Even the temperature variation as exhibited by epidote occurred as a fracture-filling material (considered as a late crystallized phase) reported an overlapping range of ~236 °C to 366 °C (except for one showing 161 °C) with the above chlorite thermometry data (~217 °C to 340 °C). For a crude comparison, the chlorite present in association with the epidote is considered here to comprehend more about the temperature of the residual fluid, although less data are available for such pairs. Three such data points associated with epidote (enclosed in biotite) suggest a much lower temperature value (~264 °C to 320 °C [69,70]; ~258 °C to 324 °C [72,73]) in contrast to the epidote temperature data (~334 °C to 529 °C), whereas in one instance, chlorite showed a similar temperature variation of ~318 °C to 321 °C as epidote (coexisting with opaque minerals), which was ~334 °C. Thus, it appears that the temperature of the late-stage fluid that prevailed within the BG may be from ~217 °C to 340 °C.

Figure 7. (a) Scatter plot of T_Ab vs. T_Or in feldspar pair [51] where T_Ab and T_Or represent the temperature for the albitic and orthoclase component, respectively; (b) a scatter plot of Ti vs. [Mg/(Mg + Fe)] for biotite with temperature isotherms [65]; (c) a scatter plot of temperature vs. pistacite content (Ps) in epidote; (d) variation in temperature in chlorite using the empirical formulation as shown in the legend [69,70,71]; (e) analyzed biotite showing oxygen fugacity between QFM and NNO [75].

Figure 7. (a) Scatter plot of T_Ab vs. T_Or in feldspar pair [51] where T_Ab and T_Or represent the temperature for the albitic and orthoclase component, respectively; (b) a scatter plot of Ti vs. [Mg/(Mg + Fe)] for biotite with temperature isotherms [65]; (c) a scatter plot of temperature vs. pistacite content (Ps) in epidote; (d) variation in temperature in chlorite using the empirical formulation as shown in the legend [69,70,71]; (e) analyzed biotite showing oxygen fugacity between QFM and NNO [75].

7.2. Oxygen Fugacity

According to Anderson and Smith [56], the oxygen fugacity indicated by mafic silicate minerals can be qualitatively comprehended by using the Fe/(Fe + Mg) ratio. The ratio depicts a decrease in value with respect to the increase in the oxygen fugacity condition. In the present case, the ratio in hornblende varies from 0.36 to 0.76 (avg = 0.52, ±σ 0.075) and is thereby indicative of variable oxygen fugacity which broadly varies from low to high during the evolution of the BG. The oxygen fugacity varied predominantly between QFM and NNO as per the formulations of Wones and Eugster [75] (Figure 7e) and was rarely above the HM buffer.

Miles et al. [76] proposed that the Mn content in apatite could be used as an indicator of oxygen fugacity in silicic magma. They inferred that the Mn content in apatite varies linearly and negatively with the log f_O2, although uncertainties involving their partitioning in apatite are there with respect to the redox state as well as other physicochemical parameters. Nevertheless, the empirical evidence associated with natural samples shows consistent results with this tool. The present data suggest a broad variation in log f_O2 values of −15.38 (+1.5) to −9.75 (+0.45) corresponding to the range of Mn concentration in apatite. Some values are higher than expected at temperatures corresponding to apatite and higher than those deduced from biotite, implying a more oxidized condition of the BG magma. The association of titanite with magnetite and quartz (assuming quartz to be a common phase in granite) in the present study suggest an f_O2 slightly higher than the QFM buffer [77]. Such a redox condition is also consistent with the redox state as revealed by the biotite of the BG, as can be seen in Figure 7e.

7.3. Volatiles in the BG

7.3.1. Water Content

A crude estimate of the liquidus temperature of the BG at 3.5 kbar varies from ~1017 to 1135 °C based on the major element data of Kaur et al. [10]. At these P–T conditions, water saturation was calculated to be 7.68 to 8.94 wt%. The formulation provided by Holtz et al. [78] for a rhyolitic melt of 800 °C was taken into consideration in order to comprehend the water solubility in the BG at its emplacement conditions. The efficiency of the model can be achieved within a pressure interval of 0.3 to 8 kbar. The pressure estimated using the equation of Schmidt [52] was used for computation of the initial water content of the BG. A value in the range of 5.25 to 10.78 wt% (except for one instance, 11.17 wt%) was obtained for the BG. The bulk geochemical data were then normalized to 100% with the estimated water saturation values using PELE [79], and the liquidus temperature was re-calculated with respect to water-saturated conditions, which suggests a value of ~754 °C to 967 °C. This temperature is very similar to the apatite saturation and zircon saturation temperatures, which are ~809 °C to 966 °C and 852 °C to 977 °C, respectively (utilizing the Kaur et al. [10] database on major element oxides of the BG). According to Puziewicz [80], there could be a temperature difference of 100 °C to 200 °C in the crystallization of an anhydrous and hydrous granitic melt. Usually, water-unsaturated granitic melt crystallizes at 900 °C to 850 °C. In this respect, the liquidus temperature as estimated using PELE with water-saturated conditions seems appropriate.

7.3.2. Halogen Content

The F/Cl value in biotite and hornblende was examined in relation to Ti in biotite and related hornblende–plagioclase geothermometers. Since these two geothermometers are the best available at this moment to understand solidus to sub-solidus conditions, accordingly, the variation in halogen concentration can be seen in Figure 8a. Despite being significantly preferred over the hornblende and biotite geothermometer for integrating both F and Cl, apatite was not considered in the current context due to the lack of a suitable number of temperature estimates for it. Furthermore, apatite is an accessory phase and comes early in the crystallization sequence for a granitic magma and results in only a slight change in the bulk halogen concentration [81,82]. As far as hornblende and biotite are concerned, hornblende has a poor preference for F, while biotite can account for up to 75% of the F budget in the granitic domain. In contrast to this mineralogical constraint for the F budget, the Cl concentration is mostly controlled by the rock’s composition [81]. Thus, the significant fluctuation in F/Cl in biotite in the BG (Figure 8a) may be attributed to the crystallization of biotite and hornblende, respectively. Even at this point, the crystallization of feldspar and quartz will raise the F content in the residual melt, partially mitigating the effect of biotite. The evolution of volatiles in biotite is susceptible to the magmatic–hydrothermal transition because of its increased sensitivity to changes in the physico-chemical environment [66,82,83,84].

According to Munoz [83], the presence of halogens in the biotite structure helps to constrain the halogen activity in the fluid in equilibrium, although the sole and direct utilization of their concentration is not recommended. This is due to the fact that the halogen content in the biotite structure is a function of temperature, the relative fugacities of the halogens in the fluid or magma, and the cation composition of the biotite, specifically the Xphl. Hence, Munoz [83] introduced halogen intercept values such as F(IV), Cl(IV), and IV(F/Cl) in order to accommodate the Fe–F and Mg–Cl avoidance rule. The IV(F) values show a nearly homogeneous distribution from 1.5 to 2.0 with very few in the range of 2.0 to 2.5 (Figure 8b). However, the IV(Cl) values of biotite show at least two different clusters (Figure 8c) primarily based on Cl% contents. The distribution of IV(F/Cl) (Figure 8d) is also consistent with Cl(IV), thereby suggesting the re-equilibration of biotite by a later fluid. It is important to note that IV(F/Cl) is invariant to the uncertainties involved in the hydroxyl occupancy and is directly associated with the relative fugacities of HF and HCl in the fluid or melt. Hence, this parameter is more dependable than that of IV(F) and IV(Cl) [85]. Since IV(F) and IV(Cl) are inversely proportional to the X_F and X_Cl values, respectively, smaller values imply a higher activity of halogen contents in the fluid [83]. In the present study, negative values of IV(Cl) and positive values of IV(F) are clearly evident in all the analyses. Hence, the higher negative values will correspond to a higher enrichment in chlorine in the fluid. Such a relationship between Cl and F could be ascribed to the higher affinity of Cl for the fluid phase than that of the solid (mineral) in comparison to F in the effect on fluid ingression [81,86].

According to Zhu and Sverjensky [66], biotite formed under similar physicochemical environments shows a linear trend in binary plots considering the parameters as shown in Figure 9a–d. Although some crude linear trend is evident in the present case, there is a dispersion of the data points due to variation in the X_Cl/X_OH and X_F/X_OH values. Hence, the composition of the fluid associated with the biotite during its formation or re-equilibration was modified by the interaction of the late-stage hydrothermal fluid, which must have a different temperature and composition. This can be depicted by the sub-solidus exchange reaction, which is common in the BG, as evident by the pronounced chloritization of biotite and saussuritization of plagioclases. In order to understand the fugacity ratios of the fluid associated with biotite, the formulations of Munoz [87] were used. Since the calculation requires a temperature value, the average temperature of the chlorite geothermometer was used in the current work and the resultant values are shown in Figure 9e,f. The log(f_HF/f_HCl)^fluid exhibits a positive linear correlation with log(f_H2O/f_HCl)^fluid (Figure 9e), whereas the plots of log(f_H2O/f_HF)^fluid vs. log(f_H2O/f_HCl)^fluid are more dispersed (Figure 9f).

8. Late-Stage Fluid Characteristics in the BG

In order to visualize the scenario of the evolution of the BG, it is essential to understand the characteristics of the late-stage fluid during its evolution. To achieve this goal, fluid inclusion studies in the matrix quartz of the BG and the late-stage fluid associated with the quartz veins/stringers were considered. A total of 24 wafers were used for the fluid inclusion study, out of which 21 samples are from the BG and 3 are representative of quartz veins.

8.1. Fluid Inclusion Petrography

Petrographic features suggest that the BG was the least affected by deformation compared to the other litho-units of the craton. As demonstrated by Tarantola et al. [88], fluid inclusions in the matrix quartz of the BG are likely to preserve their pristine compositional characteristics. Keeping the possible effect of later deformation in mind, care has been taken while choosing the fluid inclusion population for microthermometric analyses by restricting their sizes to moderately small inclusions (~5 to 15 µ) and to those with the least distortion. Thus, a reasonable assumption can be made with respect to the salinity and temperature characteristics. In the absence of cathodoluminescence (CL) images, discriminating fluid inclusion populations into discrete episodes of fluid entrapment and into different FIAs [89,90] would be misleading. Instead, we aim to comprehend the broad P–T history of the evolving BG by comparing the fluid inclusions in matrix quartz, which represents the first fluid event, with those of quartz veins. These fluids are likely to overlap with the release of fluids in discrete amounts to incipient fractures to form the veins. Therefore, in order to understand the various pulses of fluid activity covering all possible categories of fluid inclusions, features such as their occurrence in a random three-dimensional framework in a space away from healed fractures (henceforth simply referred to as “random three-dimension”) and distribution in healed fractures as trail-bound inclusions, each with the tentatively identified compositional type, were analyzed in the host quartz. Although we have concentrated on those that occur randomly in three dimensions in the host quartz and also meet the requirements for primary inclusions [91], we do not assume that they were trapped at the same time or under the same fluid P–T–X state. Even though all of these inclusions may not have been trapped simultaneously, but rather, at different times in recrystallized domains (although undeformed), the scenario with shifting fluid features in terms of temperature and salinity distribution of data is expected to furnish the spectrum of fluid activity over time rather than at any instant in time. To compare with the main population that occurs “randomly in three dimensions”, we have also taken into consideration a small representative population of inclusions in healed cracks (trail-bound).

Types of Inclusions

Based on the appearance of fluid inclusions at room temperature, along with their heating and freezing behavior, a total of six compositional types of inclusions were recognized and a summary is given below.

• Type I: Aqueous Biphase (L + V)

This type is the most dominant in the matrix quartz and vein quartz samples. They range in size from <2–30 μm (typically 5 to 15 μm) with regular (elliptical, oval rectangular, negative crystal) and irregular shapes. In the case of matrix quartz, the trail-bound nature of inclusions outnumbers the inclusions in clusters (Figure 10a). It was difficult to ascertain the primary or secondary nature of these trails due to uncertainty in the transgranular nature of the trails with other minerals in contact with quartz. This type is also seen in association with all other types of fluid inclusions.

In the case of vein quartz, the cluster mode of occurrence of Type I inclusions is dominantly seen in inclusion trails (Figure 11a). These inclusions occur in association with other types of inclusions with respect to their availability in a sample.

• Type II: Pure Carbonic

These are seen as both monophase and biphase inclusions (Figure 10b) at room temperature with sizes varying from <2 to 10 μm. They are mostly oval to elliptical in shape. Their exclusive occurrence in clusters is often observed along with an association with Type I (Figure 10j) and Type IV inclusions. In a few instances, the Type II inclusions of the matrix quartz were marked with bright white solid-like phases in the inclusion cavity (Figure 10c).

• Type III: Mixed-Pure Carbonic

In a few of the samples, the carbonic inclusions were frequently enclosed within another phase that had the appearance of a glassy type material. This material has a faint transparent to colored-translucent outline. Carbonic liquid (±vapor bubble) is either at the center or marginal (Figure 10d) to the fluid inclusion cavity. The size of such inclusions varies from <10 to 30 μm with a dominantly elliptical-to-oval and variable irregular shape. Such inclusions are also observed in one of the quartz vein samples (Figure 11d,e), where they either occur as isolated inclusions or in association with Type I inclusions. Such associations of carbonic liquid inclusions with solids are very characteristic of samples of the BG (both in the matrix quartz and veins).

• Type IV: Aqueous Carbonic

The occurrence of these inclusions is less frequent and restricted to only a few of the samples, occasionally outnumbering (matrix quartz in sample D6-12) the Type I inclusions of the BG and one sample of the vein quartz. They occur dominantly as clusters with Type I (Figure 10e), sometimes with Type II and, in rare cases, with both along the trail in the matrix quartz as well as the vein quartz (Figure 11k) of the BG. Isolated occurrences of such inclusions are also seen infrequently. They vary in size from <2 to 20 μm (mostly 5 to 10 μm) with oval-to-elliptical outlines. They occur both as biphase and triphase inclusions under room-temperature conditions. The volume fraction of the carbonic phase seems to be much higher with L_CO2 observed as a thin film. This creates problems in distinguishing the inclusions from Type I inclusions to some extent. Similar to Type II inclusions, these inclusions also contain a bright solid phase for inclusions in the matrix quartz (Figure 10f) and quartz-in-quartz domains (Figure 11i).

• Type V: Aqueous Polyphase

Inclusions containing solid phases of variable shapes and sizes such as cubic, rhombohedral, prismatic, globular, and occasionally trigonal and columnar in shape are seen in the matrix quartz (Figure 10h). The solid crystals are dominantly prismatic in the vein quartz in the BG (Figure 11g,h). The inclusion size varies from <5 μm to 20 μm in both cases. Solid phases occupying more than half of the fluid inclusion cavity are seen in both the domains as shown in Figure 10i and Figure 11l,m. These inclusions contain more than one solid phase, indicating that these solids may be accidentally trapped. However, their recurrence is quite common in fluid inclusion assemblages.

• Type VI: Unusual Inclusions

This type is more or less similar to the Type V inclusions described earlier in the “giant quartz reef system” [4], but they are less frequent in samples from the BG. This type has single or multiple tiny entities executing pseudo-Brownian movement, often giving an effervescence-type behavior and are quite few in number rather than the many variations in Type VI inclusions that are observed—the movement of the suspended particles is feeble and observed only after a slight increase in temperature, making it difficult to distinguish these inclusions from Type V inclusions. Moreover, inclusions exhibiting a colored aqueous liquid (viz. brown or green color) or sometimes colored phases in the fluid inclusion cavity are comparatively more frequent (Figure 10i) than those of the effervescence-type of Rout et al. [4]. Although these inclusions behave as Type I inclusions, they were categorized under the unusual type in the current work. They are seen in association with Type I inclusions in both clusters and trails. The isolated occurrence of these inclusions is observed occasionally. Type VI inclusions are relatively less frequent in the quartz veins (Figure 11n). Occasionally, the aqueous liquid is colored as shown in Figure 11f.

8.2. Fluid Inclusion Microthermometry

A total of 543 fluid inclusions in the matrix quartz of the BG and 83 fluid inclusions from its associated quartz veins have been studied and their details are given below.

The majority of inclusions in both the matrix and vein quartz in the BG are Type I inclusions. The temperature of first melting of inclusions in the matrix quartz shows a broad range from −53 °C to −21 °C indicating the presence of cations like Mg, Ca, and K along with Na in the aqueous fluid. The salinity values obtained from their final ice melting temperatures range from ~0 to 26.85 wt% NaCl eq. Such variations in salinity values are consistent in both the trail-bound and non-trail-bound inclusions (Figure 12a). There is also a dominance of low-salinity (0 to 2 wt% NaCl eq.) inclusions compared with moderate-to-higher salinity values. A similar situation prevails in Type I inclusions of the vein quartz hosted in the BG, in which salinity values vary from ~0 to 21.06 wt% NaCl eq. but cluster around 0 to 2 wt% NaCl eq. The temperature of homogenization varies from 94.8 °C to 399.8 °C (clustering around 120 °C to 300 °C) and 73.2 °C to 357.3 °C (clustering around 130 °C to 280 °C) in non-trail-bound and trail-bound inclusions, respectively (Figure 12b). This results in a density distribution of 0.507 to 1.136 g/cc in non-trail-bound and 0.602 to 1.177 g/cc in trail-bound Type I inclusions in the matrix quartz. Even the broad range of temperature of homogenization in the vein quartz is marked and varies from 139 °C to 360 °C with comparatively narrow clusters around 150 °C to 220 °C.

The density values of these inclusions range from 0.57 to 1.069 g/cc (Figure 12c). All Type I inclusions homogenized into the liquid phase with the exception of one inclusion exhibiting critical homogenization. The T_m,CO2 value in Type II inclusions is consistently close to the triple point of CO₂ (−56.6 °C), indicating undetectable concentrations of other gaseous species. The homogenization temperatures of these inclusions (T_h,CO2) span a broad range from −7.6 °C to 31 °C with clusters around 25 °C to 30 °C (Figure 12d). The density of these inclusions varies from 0.513 to 0.971 g/cc and clusters around 0.6 to 0.8 g/cc (Figure 12e). However, T_h,CO2 in Type III inclusions shows a narrow range from 21.2 °C to 30.2 °C with the density of the carbonic phase varying from 0.586 to 0.761g/cc. In both cases, homogenization is into the liquid phase. The bright phases present within the carbonic liquid do not show any tendency for dissolution on heating, possibly due to extremely slow kinetics, since they are not chloride salts. These solid phases in the carbonic inclusions are more likely to be captive phases, although the occurrence of such inclusions is quite frequent. The similar near-pure nature of CO₂ is also marked in Type IV inclusions in both the matrix and vein quartz in the BG (Figure 12d). The melting temperature of clathrate in these inclusions varies from −7.1 °C to 6.9 °C. The temperature of partial homogenization of the carbonic phase (T_h,CO2) ranges from 17.3 °C to 30.7 °C, with the mode of homogenization being mostly to the liquid phase but sometimes to the critical state. The temperature of total homogenization (T_Tot) could only be recorded for four of the Type IV inclusions. The rest of them decrepitated before total homogenization. Out of the four Type IV inclusions which furnished T_Tot, two homogenized to the aqueous phase at 328.7 °C and 356.2 °C; one into the carbonic phase at 399.3 °C; and another one was on the verge of homogenization to the carbonic phase at around 600 °C (since the equipment could not go beyond 600 °C). Out of these four, T_m,Clath could only be recorded precisely for two inclusions, i.e., the one that homogenized at 329 °C and another at 600 °C. Since the complete homogenization could not be noted in the latter case, only a single Type IV inclusion temperature could be used for barometry (estimation of minimum pressure). The clathrate melting temperature gave a salinity of 9.84 wt% NaCl eq. A pressure of 2.6 kbar was obtained using the MRK equation of state [92,93] following the computation scheme outlined by Panigrahi and Mookherjee [94]. This inclusion had a calculated density of 0.97 g/cc, having an X_CO2 of 0.07 at the temperature of total homogenization. It may be noted that this pressure corresponds to the minimum pressure of entrapment. However, since some of the decrepitated inclusions were also approaching homogenization to the carbonic phase, entrapment close to the solvus in the approximated ternary system seems possible, and hence, the pressure estimated could be close to the pressure of entrapment.

None of the Type V inclusions exhibited the dissolution of solid phases present within the inclusion cavity upon heating. However, the disappearance of the vapor bubble took place in a temperature interval of about 131.7–330.5 °C. No distinct behavior or phase changes could be observed for the vibrating phase present in Type VI inclusions upon freezing except that its movement is arrested at low temperature. Increasing the temperature causes the movement to become relatively vigorous, making it difficult to visualize the vapor bubble disappearance. Only in one instance did the random movement of the solid phases stop in response to heating, which occurred at 429 °C. In other cases, partial homogenization was achieved at temperatures of 177.4 °C to 296.8 °C. The density calculated from the partial homogenization temperatures varies from 0.713 to 0.9 g/cc.

8.3. Laser Raman Microspectrometry

Laser Raman microspectrometry (LRM) of some of the Type I inclusions reports a water peak with a broader hump which indicates the low-salinity nature of the aqueous fluid in these inclusions [95,96,97]. The occurrence of various compositional types of fluid inclusions containing various solid phases and different colorations of the aqueous liquid does not change during microthermometry, Hence, LRM was carried out to characterize the usual and unusual inclusion types in the BG domain. In reference to the occurrence of solid-like phases within the inclusion cavity as revealed from fluid inclusion petrography, the focus was given to the Type III and Type V inclusions. LRM identified many silicate minerals such as mica/biotite, albite/anorthoclase, and titanite along with carbonate and hydroxides such as calcite and edenite within the inclusions in the matrix quartz in the BG (Figure 13a–e). The carbonic phase in Type II and IV is devoid of any peaks other than CO₂ as shown in Figure 14a,b The presence of calcite in Type IV inclusions is indicated from its Raman peak as shown in Figure 14a. LRM of inclusions in the quartz vein samples also indicates the presence of silicate minerals such as pectolite and anorthoclase and the possible identification by LRM of carbonate minerals like burbankite [(Na, Ca)₃(Sr, Ba, Ce)₃(CO₃)₅]; it is an accessory mineral in carbonatite, although this peak could alternativelybe an unidentified organic compound. The identification of pseudobrookite (an iron titanium oxide, Fe₂TiO₅) is shown in Figure 15c. Rarely, the presence of a glassy phase is also detected within the aqueous part of the inclusion cavity (Figure 15b). One of the Type VI inclusions present in vein quartz (Figure 11n) was considered for study but did not show any characteristic Raman peaks except for a broad water band (Figure 15d). Some representative Raman peaks and maps of fluid inclusions in vein quartz are also given in Figure 16.

8.4. Fluid Evolution

Attempts have been made to retrieve the trapping temperature and pressure from the coexisting Type I and Type II inclusions assuming they are coeval. However, only two such pairs could be identified in the matrix quartz in the BG, and the P–T values as determined from the intersecting isochores method [98] are 220 °C/850 bars and 270 °C/1560 bars (Figure 12f). Such an estimation for P–T corresponds to the immiscible region in the H₂O–CO₂–NaCl system and depicts a fluctuating fluid pressure, possibly between purely lithostatic to intermediate conditions (~3 to 4 km deep), whereas a Type IV inclusion furnishes a pressure of 2.6 kbar, which corresponds to a depth of about 8 km and the miscible regime of the aqueous–carbonic fluid. Moreover, the above pressure value (2.6 kbar) is well correlated with the minimum pressure estimation from the hornblende barometry. These Type IV inclusions possibly represent the earliest stage of fluid exsolution from the BG which exhumed to a shallower depth (indicated from pressures obtained from the intersecting isochores) and underwent brittle deformation.

The distributions of data points in the temperature versus salinity plot (Figure 17) indicates a dominant mixing trend parallel to the temperature axis. It may be noted that such a fluid evolution diagram remains incomplete since data of the carbonic and aqueous–carbonic inclusions cannot be included on such a plot. In addition to this, aqueous–carbonic inclusions represent the higher temperature range in the thermal spectrum. In addition, temperatures of homogenization are not pressure-corrected and only give a tentative picture. Within these limitations, a mixing trend with two fluid components—one a low-salinity and low-temperature meteoric fluid and the other a moderate-salinity and higher-temperature fluid component—could be identified. The consistency in the trend can also be marked in all domains as shown in the legend. Moreover, it is also observed that the low-salinity (~5 wt% NaCl eq.) fluid has a considerable spread in temperature up to ~400 °C, possibly ascribed to transient heating of the fluid during its evolution in the BG. Alternatively, this high-temperature fluid of low salinity could also represent the earliest phase of exsolved fluid from the BG magma. Thus, a close look at the fluid evolution diagram (Figure 17) indicates the interplay of three different fluids: (i) a low-temperature–low-salinity (LTLS) fluid, (ii) a high-temperature–low-salinity (HTLS) fluid, (iii) a moderate-temperature and moderate-to-high-salinity (MTMS) fluid. The LTLS fluid in all probability is a meteoric fluid. The MTMS fluid in the BG is interpreted to be the late-stage fluid that evolved internally within the BG by the incursion of meteoric water. The HTLS fluid with CO₂ in the BG is a definitive magmatic component and may represent an early fraction of the fluid exsolved from the granitic magma with the initiation of quartz crystallization. The low salinity is expected since a higher concentration of CO₂ would lower the solubility of salts. The immiscibility is also evident by the association of Type I, II, and IV inclusions, though they are rare or restricted in occurrence. In such instances, there is a greater possibility that the partitioning of the solutes present in aqueous–carbonic fluid will be in its aqueous component rather than its carbonic counterpart. This, in turn, can also increase the salinity in the aqueous fluid.

8.5. Significance of Type V and VI Inclusions

Microthermometry conducted on the solid-phase-bearing inclusions as well as the unusual type inclusions failed to provide information related to temperature as well as salinity. Raman spectra of many such solid phases suggest that they are silicate and carbonate minerals or rarely oxide. One possible origin for these phases could be their captive nature indicating their presence as suspended mineral particles in the fluid precipitating quartz, while the fluid was evolving in a closed system. The captive phases in these inclusions (Type III and VI) are not different from the secondary mineral phases in the BG; hence, they are explainable. In contrast to this, the recurrence of such phases in fluid inclusion assemblages could also suggest the possible presence of a silicothermal fluid (K₂O–CO₂–SiO₂–H₂O system as suggested by Wilkinson et al. [99] from which they might have precipitated as daughter crystals during the decrease in P–T conditions. According to Wilkinson et al. [99], a silica-rich (~90 wt% SiO₂) fluid can coexist with a supercritical, alkaline, aqueous–carbonic fluid in which immiscibility can be achieved in P–T conditions of <200 MPa and 300 °C to >750 °C, respectively. In the present context, the Type III inclusions in the matrix quartz in the BG is very similar to the inclusions reported by them. These inclusions are also seen to coexist with Type I (aqueous biphase) and Type IV inclusions (aqueous–carbonic). Thus, the aforementioned assemblage could owe its origin to heterogeneous entrapment, suggesting immiscibility in the late-stage fluid that evolved in the BG. Many workers also inferred the occurrence of siliceous fluid along with aqueous fluid at variable P–T conditions, such as Morey and Fleischer [100] who suggested a moderate temperature (340 °C–500 °C) and a low pressure value of 20–40 MPa in the system K₂O–SiO₂–H₂O. If, at all, they represent the above KCSH system, they could be vital in reference to the huge mobilization of silica in such shallow crustal domains without the demand for a staggering amount of fluid.

Another possibility could be that they might represent the primary silicate gel as reported by Thomas and Davidson [101]. According to them, gel inclusions have a resemblance with the fluid inclusions with a shrinkage bubble, the larger volume fraction (50 to 100%) of which is occupied with solid materials. Essentially, they can be visualized as micrometer-sized granular masses, primarily quartz grains which may form irregular clusters after the accretion of the aforementioned granular masses. In addition to feldspar, FeOOH and cristobalite occur within it, leaving the remaining portion of the fluid inclusion cavity occupied by an alkali-carbonate-bearing aqueous solution and a vapor bubble. Thus, quartz, feldspar, and calcite constitute some common solid phases in them. Likewise, the BG also has some notable occurrences of calcite-bearing aqueous–carbonic inclusions along with other silicate phases like albite. Thomas and Davidson [101] suggested their probable formation in a colloidal state which may act as a useful link for the formation of giant quartz reefs in the BC.

9. Discussion

The dominance of the very coarse to medium-grained nature of the BG suggests its emplacement in a range of depths from deeper to intermediate levels in the crust and this is supported by the pressure data obtained using the hornblende barometer [52], that is >1 kilobar to ~6 kilobar. Since hornblende is not uniformly abundant in the BG, a clear picture of the special variation in pressure could not be obtained, which would have otherwise indicated different crustal levels of exposure of the pluton. In all, the BG originated at great depth (~18 km/6 kbar) from which the crystallization was initiated progressively through mid-to-shallow crustal (~8–10 km/1–2 kbar) conditions. During this long stretch, the crystallization history from the magmatic temperature to its sub-solidus condition can be retrieved through various geothermometers that depicts various stages of its evolutionary history. According to Naney [102], magmatic biotite most likely crystallizes or equilibrates at approximately 680 °C (near-solidus temperature) and remains stable up to the complete crystallization of the rock. Thus, the three different temperature ranges of apatite–biotite thermometry in a sample may corroborate three different events of thermal history in the BG during its evolution. The higher temperature (~823 °C) may be indicative of a minimum magmatic temperature for the BG. This is due to the fact that apatite comes first in the crystallizing sequence in felsic magma, around approximately 900 °C–924 °C [103,104], thereby establishing the equilibrium relationship with the enclosing biotite and said temperature seems plausible. Since all of the biotite analyses are plotted in the field of re-equilibrated biotite [32], mostly close to the primary biotite field, the Ti thermometry in them may suggest re-equilibration near to the solidus after the solidification of the BG. This coincides with the median temperature range of the apatite–biotite geothermometer (671 °C–685 °C). Thus, the lower thermal signature (393.6 °C–489.6 °C) of the apatite–biotite pair implies that there might have been some later modification to the equilibration in response to the later fluid activity [85], although this geothermometer is less prone to sub-solidus re-equilibration compared to carbonates and oxides [67].

Combining the present petrographic study and mineral chemical analysis of epidote, some fully or partially enclosed epidote within biotite and hornblende, associated with felsic minerals, such as plagioclase with resorbed outlines, can be presumed to be of magmatic origin. However, there is some difference of opinion [45,47,105] in establishing its magmatic nature. It implies that the BG must have experienced a rapid ascent to avoid the complete resorption of mEp. Hence, diking is a more appropriate mechanism for their emplacement rather than diapirism [106]. The appearance of mEp in a granitic magma at moderate-to-higher pressure (6 to 8 kbar) is very much dependent on the composition and emplacement conditions of the granitic magma [45,102]. The experimental study of Schmidt and Thompson [107] on epidote in calc-alkaline magma at H₂O saturation and an f_O2 of NNO buffer conditions depicts a minimum of 5 kbar pressure within the stability field of epidote in a tonalitic magma. At higher oxygen fugacity conditions such as in the HM buffer, the stability field is widened up to 3 kbar. The resemblance of such pressure ranges to the present hornblende barometry data and the broad oxygen fugacity supports mEp transport in the BG.

Our observations and data do not corroborate the hypothesis of the interplay of hot (~900 °C) mantle-derived fluid resulting in the filling of a mesoscale fracture network in pink K-feldspar megacrysts with melanocratic veins, as proposed by Sensarma et al. [108], based on the high FeO content of the K-feldspar. The K-feldspar megacrysts preserve their triclinicity and are unlikely to have been infiltrated by such high-temperature fluids. According to Teiber et al. [81] and Słaby and Domańska-Siuda [82], mantle supplements are usually rich in Cl and the interaction of even a small volume of mantle-derived material will enhance the Cl concentration in the system. However, there is no significant variation in F/Cl ratio in the hornblende (Figure 8a). Although, some fluctuations in F/Cl ratios were seen in biotite, they were not substantial and also indicated an increase in F/Cl which should have been the reverse if they would have interacted at such a high temperature (950 °C) for mantle-derived melt/fluids. Moreover, such reddening is a usual feature seen in Proterozoic granites worldwide such as granitic terranes in the Bastar and Eastern Dharwar Cartons [94,109] where pink granites are ubiquitous and localized reddening of K-feldspar is observed wherever there is intense fluid activity (often with mineralization) without any spatial influence of mafic dykes. Putnis et al. [110] and Plümper and Putnis [111] also supported the occurrence of sub-microscopic inclusions of hematite in the alkali feldspar to be the main reason behind reddening in such Proterozoic granites, as did Sensarma et al. [108]. However, according to them, hematite inclusions in feldspar could have variable origins such as co-precipitation from granitic melt, they could be the result of solid-state exsolution from an Fe³⁺-bearing solid solution, or they could be the manifestation of the external introduction of Fe-bearing fluids to the crystallizing granite. It is also considered that sub-solidus deuteric reactions in the presence of fluid are another factor which causes brick-red coloration in Pre-Cambrian granites as suggested by Plümper and Putnis [111]. Further, Colleps et al. [112] reported the long-term low-temperature evolution of the ~2.5–3.3 Ga Bundelkhand Craton and the surrounding ~0.9–1.7 Ga Vindhyan basin in central India. Petrography of the BG also supports a pervasive alteration in the BG rather than any structure-controlled alteration, which is well marked with variable degrees of sericitization of plagioclase and the chloritization of hornblende and biotite. The occurrence of specks of magnetite within hornblende as a distinct rim as well as fine-grained quartz and titanite (petrography section) can be inferred here to indicate the deuteric alteration of hornblende, possibly in a more oxidizing fluid. The similar and wide range of Ps content in most of the textural categories of the epidote in the current study might be due to a broad disparity in f_O2 conditions in the system [45,47].

The estimation of water content using the formulation of Holtz et al. [78] furnished variable amounts of water saturation (5.25 to 10.78wt %) for the BG during its evolution with hornblende as a crystallizing phase. The importance of the initial water content on the occurrence of equilibrium assemblages in magma can be furnished from the experimental work of Naney [102]. He proposed a paragenetic sequence of orthopyroxene, biotite, clinopyroxene, orthopyroxene (resorbed), plagioclase, alkali feldspar, clinopyroxene (resorbed), and quartz for a granitic magma at 5 wt% H₂O and 2 kb pressure. On the other hand, he inferred that an augitic clinopyroxene crystallized from an anhydrous bulk granitic composition replaced by hornblende when changed to a granodioritic composition. This might be another alternate perspective upon the commencement of normative pyroxene for the BG in CIPW normative calculations without any modal pyroxene or olivine in them, whereas Sensarma et al. [13] linked this normative pyroxene to the green mafic veins of variable scales dispersed within the BG. On the other hand, the presence of hornblende in a granodioritic system suggests a minimum of 4 to 2.5 wt% H₂O content for its stability at the pressure value of 2 to 8 kbar, respectively.

Fluid inclusion data are reported with the dominance of low-salinity fluid (0–5 wt% NaCl equivalent) in the matrix as well as in the vein quartz, which is well observable within a broad span of temperature of ~100 °C to 400 °C. The overall fluid characteristics as depicted in the fluid evolution diagram provide a qualitative picture for the evolution of fluid in the BG. Although, high-temperature CO₂-bearing fluids are reported from the fluid inclusion study, they do not compare well with the metamorphic fluid as reported in the granite greenstone terrane [113]. Although a carbonic fluid is present, its restricted presence in some of the samples as well as the lack of CH₄ favors a magmatic fluid over a metamorphic one. In addition, the prevailing action of meteoric fluid, which can be visualized as a mixing trend in the fluid evolution diagram, suggests an internal evolution of fluid within the BG domain. Hence, the present work mostly visualizes an internally evolved fluid within the BG instead of a pristine magmatic fluid [94,114,115]. While Allan and Yardley [114] ascribed the incursion of meteoric fluid for the pervasive deuteric alteration in the granitoids, Reed et al. [115] demonstrated that a single magmatic fluid can also evolve to LSLT condition after a prolonged fluid–rock interaction. In either of these cases, fluid of such moderate temperature–salinity characteristics is quite likely to have been sourced from a granitoid, representing a late-stage pervasive fluid in intergranular spaces or a fluid resulting from the mixing of a late-magmatic fluid with a meteoric fluid. The occurrence of almost every kind of inclusion in both the matrix quartz and the quartz veins/stringer of the BG, in conjunction with the involvement of a moderate salinity–moderate temperature (MSMT) fluid, strengthens the hypothesis that similar fluid types were involved during their formation. Moreover, such a similarity can be achieved when a comparison is made with the fluid characteristics of quartz reefs and associated quartz veins [4] with the present fluid inclusion work in the matrix quartz of the BG (host of the quartz reefs) as well as the associated quartz veins. Rout et al. [4] also reported the presence of similar varieties of fluid inclusion types as in the BG regime except for Type III inclusions which are rare in the quartz reef domain.

Table 4. Comparative statistics of fluid inclusion characteristics between the giant quartz reef BG domains (3D Array—inclusions as part of a random three-dimensional network).

Statistical Parameters	Salinity				Temperature
	Quartz Reef		BG Domain		Quartz Reef		BG Domain
	3D Array	Trail	3D Array	Trail	3D Array	Trail	3D Array	Trail
Min	0	0.18	0	0	53.4	58.6	94.8	73.2
Max	28.9	24.39	26.85	29.22	385.9	326.4	399.8	357.3
Mean	6.50	9.16	5.50	6.51	191.56	182.20	211.63	203.42
Median	0.18	0.18	0.18	2.89	200.8	157.9	150.2	175
Mode	1.39	9.25	3.53	3.53	187.05	177.35	202.65	195.4
Standard Deviation	8.18	8.27	5.69	6.89	50.48	62.77	59.74	53.86

is a summary of descriptive statistics of fluid inclusion data from the giant quartz reefs and the BG (matrix quartz and subordinate veins). Furthermore, the fluid evolution diagrams have a striking resemblance with each other as if a mirror image in the first instance. Nevertheless, the disparities among them such as the prevalence of effervescence inclusions in the quartz reef domain and the dominance of Type III inclusions along with comparatively more carbonic inclusions in the BG regime are easily explainable in reference to the ambient physico-chemical environment of evolution. Thus, the precipitation of the matrix quartz of the BG, as well as the associated quartz veins and the origin of quartz reefs as well as the subordinate quartz veins, all represents the same fluid circulating within the BG and the meteoric incursion, which is continuously evolving with the interaction of the BG. They all owe their formation to the protracted period of crystallization of the BG.

10. Conclusions

The Bundelkhand granite, although comparable to other late-Archean/early-Proterozoic granitoids in the Indian shield, is unique in hosting the giant quartz reef system. It preserves imprints of crystallization evolution from magmatic temperatures and pressures (in excess of 900 °C and mid-crustal pressures) down to low-temperature deuteric hydrothermal stages. The mineral assemblage comprises zircon, apatite, hornblende, plagioclase, magmatic epidote through biotite and late epidote, chlorite, and ubiquitous quartz and K-feldspar. This gradual decrease in pressure–temperature conditions is in general agreement with the fluid evolution down to low temperatures and pressures registered in fluid inclusions in quartz. Fluid inclusion data in conjunction with petrography and mineral chemistry of the magmatic and hydrothermal phases convincingly indicate the prolonged fluid–rock interaction and the internal evolution of fluid with incursion of meteoric fluid. The almost indistinguishable fluid characteristics and fluid evolution paths in the BG and the quartz reef domains is remarkable. This leads us to visualize a “concomitant leakage” of the pervasive fluid from the BG to the major fracture system giving rise to the giant quartz reefs in a continuum. Admittedly, the proposition (as depicted in Figure 13b of [4]) does not strictly conform to the temporal relationship deduced from radiometric dating attempted by many workers as discussed in the Introduction Section. We do not intend to offer any commentary on the existing geochronologic data and would like to conclude that the fluid inclusion data are too compelling to ignore the possibility of a genetic link between the BG and the “giant quartz reef” system. The only other possible alternative to explain the origin of the giant quartz reefs is structurally controlled fluid activity post-BG that could account for the suspected time gap. However, the identical characteristics and path of evolution of fluids in the domain would be difficult to reconcile. We suggest that a more thorough investigation on zircons from both domains needs to be carried out to resolve the issue of the temporal relationship between the two.